Movies In Cisco allows you to visualize on a map all the movies which have ever been filmed in the beautiful city of San Francisco, CA, USA.
The entire app is based on an open data set made available by the city of San Francisco whose link is the following:
https://data.sfgov.org/Culture-and-Recreation/Film-Locations-in-San-Francisco/yitu-d5am
Starting from this data set which contains, amongst other informations; the location in which a scene of a certain movie was filmed.
The location is not indicated as coordinates but just as an address.
This forced me to convert every address for every entry in the data set to coordinates using Google's geocoding API:
https://developers.google.com/maps/documentation/geocoding/overview
The problem with this particular API is that it is restricted to a maximum number of requests per second (50), so i had to configure multiple threads fetch data quickly but looking out for a potential query limit error generated by the web-service, the problem still needs to be solved but with an average speed connection it shouldn't occur.
The converted locations are then saved in a local database, to get around the conversion effort every time the app is restarted, and shown on the map as clusters, for performance related reasons.
Once a user zooms in, the clusters divide in smaller clusters eventually revealing a marker which can be clicked to show the movie filmed in that location and its poster.
If multiple movies where filmed in the same location a cluster will be available to click generating a popup window showing all the different movies.
The application is able to load posters for most of the movies thanks to the TMDB API:
https://developers.themoviedb.org/3/getting-started/introduction
Given the name of a movie the web-service responds with a set of information about it, including the path of the poster on the TMDB's database, which can then be retrieved and loaded on the app thanks to the Glide framework. This part of the app is a bit rusty since I couldn't put the request function in a separate thread resulting in the UI locking sometimes until the request is completed.